Classification between normal and abnormal respiratory sounds based on maximum likelihood approach

Abstract

In this paper, we have proposed a novel classification procedure for distinguishing between normal respiratory and abnormal respiratory sounds based on a maximum likelihood approach using hidden Markov models. We have assumed that each inspiratory/expiratory period consists of a time sequence of characteristic acoustic segments. The classification procedure detects the segment sequence with the highest likelihood and yields the classification result. We have proposed two elaborate acoustic modeling methods: one method is individual modeling for adventitious sound periods and for breath sound periods for the detection of abnormal respiratory sounds, and the other is a microphone-dependent modeling method for the detection of normal respiratory sounds. Classification experiments conducted using the former method revealed that this method demonstrated an increase of 19.1% in its recall rate of abnormal respiratory sounds as compared with the recall rate of a baseline method. It has also been revealed that the latter modeling method demonstrates an increase in its recall rate for the detection of not only normal respiratory sounds but also for abnormal respiratory sounds. These experimental results have confirmed the validity of our proposed classification procedure.

Construction of Ergodic GMM-HMMs for Classification between Healthy Individuals and Patients Suffering from Pulmonary Disease

“…In our previous studies [13]- [18], we constructed a left-toright GMM-HMM with limited states for each segment, as shown in Fig. 5 (a), and assumed that the models were not suitable.…”

“…and selected the frame length and frame intervals for analysis. In our previous studies [13]- [18], we set the analysis frame length to 25 ms and frame interval to 10 ms. In this study, we selected several combinations of frame length and frame interval, as presented in Table 5.…”

“…To overcome these issues, in our previous studies, the features should be expressed statistically. The time series of the acoustic features of lung sounds are expressed by constructing Gaussian mixture model (GMM)-hidden Markov models (HMMs) to discriminate between normal and abnormal respiratory sounds [13]- [18]. However, we did not consider the suitable state transition of the GMM-HMMs, analysis frame length, and frame intervals.…”

Digital Pulmonology Practice with Phonopulmography Leveraging Artificial Intelligence: Future Perspectives Using Dual Microwave Acoustic Sensing and Imaging

“…Various AI models have been used in detecting and analyzing lung sounds [ 38 , 39 , 40 , 41 , 42 , 43 , 44 , 45 , 46 , 47 , 48 , 49 , 50 , 51 , 52 , 53 , 54 , 55 , 56 , 57 , 58 , 59 , 60 , 61 , 62 , 63 , 64 , 65 , 66 , 67 , 68 , 69 , 70 , 71 , 72 , 73 , 74 , 75 , 76 , 77 , 78 , 79 , 80 , 81 ]. These methods have been tested in and proposed to be used in a multitude of clinical settings.…”

“…Along with its use in the diagnosis and management of various pulmonary disorders including pneumonia, COPD, asthma and IPF, its ability to filter cardiovascular sounds makes it superior to conventional stethoscopes [ 76 ]. Despite extensive research in the field, the lack of substantial sample sizes and the inability of current models to filter environmental noise have hindered AI’s development and use in everyday clinical practice [ 56 ]. Furthermore, since clinical decision making is feasible with interpretable AI, current reviews of AI models are mostly black box-type models without clarity regarding the exploitability of the features that relate to underlying pathophysiology in order to guide practice.…”